This work is concerned with the problem of planning trajectories for a group of mobile robots operating in a obstacle-filled environment. The objective of the robots is to cooperatively reach targets sets, avoiding collisions between themselves and with obstacles, while maintaining the robust connectivity of their communication network. Model predictive control with mixed-integer encoding is used to plan the trajectories in a centralized fashion. Two novel constraints are introduced to cope with the robustness connectivity problem. Simulations are provided to evaluate the proposal.